This stuff is fun lately, I've been working on an engine for a while and ported a little game over to the browser here. It's a voxel kart racer as a browser project. Think classic kart racing, but with a voxel aesthetic. It's a fun little experiment, so I figured I'd share it here!
Thanks! Still needs a bit of work and polish, of course, but glad you enjoyed it! It includes multiplayer as well, which I've done a bit of testing on. If you notice anything or have any suggestions, feel free to let me know. I enjoy tinkering with these kinds of projects!
In my opinion, I think might have to do more with the dark patterns that are created by these apps. DoorDash, Uber. You place an order and there's all these hidden things that get added in you don't really see. You don't really get to recognize unless you do it manually. It's it's trickery in a way. They they try to gouge and get extra fees and costs in there where you can find ways around it if you're clever and able to do so. So using an agent, that just gives them an advantage because then you end up paying the overinflated price and fees plus tokens on top of everything else. See what people are looking for nowadays is actually ways to lower costs. Inflation is too high, and I think people are starting to realize that. They're trying to find ways and find cheaper ways to do things.
Same, I've been working on some little projects and 3d environments, its quite enjoyable overall. This one is a procedurally driven voxel game - still a work in progress, but its turning out better than expected.
Why does it quote 37,000 acres, then show a picture with a large swath of Sweden highlighted as if that is the area purchased - though Sweden is approximately 111 million acres of land, so this would be 0.003% of the land, so nowhere near the representation in this picture - seem disingenuous to me.
I think someday we’ll all have our own custom LLMs, built around who we are, what we know, how we think, and the information we’ve accumulated. That kind of contextual communication could be far more useful than simply having a better autocomplete.
I do appreciate old things, antiquity, retro technology, and nostalgia. But I also think people sometimes hold onto things long after their practical purpose has passed. That’s not necessarily bad. Typewriters, handwriting, calligraphy, old forms of art and writing can still have value as crafts and forms of expression. They just don’t necessarily need to remain the primary way we communicate.
The more interesting problem to me is communication itself. We often assume that because two people speak the same language, they’re actually transferring information effectively. In reality, two people can hear the exact same words and walk away with completely different understandings because their knowledge, context, and mental models are different.
I think the future should be less about preserving old methods of expression and more about finding better ways to actually transfer meaning between people, while still allowing each person to understand that information in their own context.
The old ways will probably stick around as art, nostalgia, and tradition, just like old forms of lore and craftsmanship. That’s fine. They’re worth keeping for those reasons. But I don’t think we need to confuse preserving the art with preserving the limitations of the old method.
So what do you do when your government is effectively bankrupt? You’ve accumulated more debt than you can realistically repay without imposing enormous economic pain, but you still need to keep the country functioning. One historical answer is an inflationary cycle combined with a trade war.
You impose tariffs on foreign goods, which creates government revenue while also making imported goods more expensive. You then blame the resulting price increases on foreign competitors and trading partners. But the tariffs themselves are inflationary because businesses ultimately pass much of those higher costs on to consumers.
As prices rise, nominal tax revenues also rise because governments are collecting taxes on larger dollar amounts. Meanwhile, governments have an incentive to continue borrowing, spending, and potentially expanding the money supply to deal with their fiscal problems. More money chasing higher-priced goods creates further inflation.
The critical part is what inflation does to existing debt. If you owe $1 trillion in debt today, but inflation dramatically reduces the purchasing power of the dollar over time, that $1 trillion becomes much easier to repay in real terms. The nominal debt remains the same, but its real economic burden shrinks.
So you can end up with a feedback loop: tariffs increase prices, higher prices increase nominal tax receipts, fiscal pressure encourages more borrowing and monetary expansion, monetary expansion fuels more inflation, and inflation gradually erodes the real value of the government's existing debt.
In other words, rather than explicitly defaulting on the debt, you can effectively "inflate it away." The creditors get paid back in dollars, but those dollars are worth substantially less than the dollars they originally lent.
The danger is that this isn't free money. The cost is ultimately transferred to consumers, savers, wage earners, and holders of government debt through reduced purchasing power. And once inflation expectations become entrenched, controlling the cycle can become extremely difficult without causing a recession or other severe economic disruption.
Hey, thanks for this comment. You seem to be knowledgeable about the topic and I am a moron.
What happens to consumer spending and associated tax revenue when imported goods cost significantly more? Let's take a hypothetical German auto brand that has no factories in the US (BMW, MB, VW have factories, not sure if Audi and Porsche are grandfathered in because of VW).
If a car that cost $100K before and now is $125K, it seems that the demographic that could afford that $100K will shrug and pay the higher price. Now, that's the well-off folks, a bit immune in this economy.
What other consumer goods with imposed new tariffs will cause less spending and shrink the economy instead? It seems that tariffs are a net loss, besides I guess in cases when they are used to prevent dumping.
Anthropic is a bit nuts, I had $260 of credits on my max account for the extra usage the other night. It was expiring, so I figured I'll fire up an agentic swarm to deep dive and make some deep changes to some old cold bases.. literally 25 minutes or less, $260 burnt, it didn't get get into the implementation, just wrote a ton of useless plans for the most part. It really opened my eyes to what they expect to charge people.. wayyyy overpriced.
Hint: The new models are really good at burning tokens.
I've had to use it a bit for work, and it's been remarkable watching the degradation in performance with the default suggested current models (Opus 5 as a prime example) vs the models that got them huge attention a year ago (Opus 4.6)
If you give 4.6 a spec, or existing code to implement a feature in, it will ask some pointed questions if there's something unclear in the spec, and then produce a plan and move to implement it.
5 will freak out at even a basic task, ask itself if it's own assumptions or your instructions are correct, proceed to re-assess it's own plan, and it's instructions 3-4 times, and then maybe produce code after burning several hundred thousand tokens (and quite a bit of time) analyzing existing code and thoroughly sweeping it for irrelevant problems both to the task it was given and the spec it came up with.
It's quite bizarre to me how well advertised the benchmarks and anecdotes from people one shotting MVP browser games are, compared to the experience of everyone I know that's had to actually use it to accomplish even a relatively basic task.
The entire ecosystem of CC is designed to facilitate burning tokens. You have to ask the LLM to write a script for the app to tell you which folder you're working in and which branch you're on. There are commands that just diagnose your Claude Code setup and try to "optimize" it. Adding skills or plugins bloats the context window. Developing plans means that you work through questions before you get to it in the code, but that matters way more for human programmers than LLMs, so it's probably just a waste of tokens.
I managed to lose around $300 in credits I had saved for some emergency /fast sessions the following way: switch to Fable. Work on the design. Downgrade to Opus for the build. If any of other parallel Opus session has /fast enabled it seems to enable it for the newly spawned session by default. Before I knew it, the $300 was gone. I think the bug is now solved, but it was rather unpleasant. I dont ever remember bugs that would drain my wallet - with claude code its just another Tuesday. Still love it.
Claude code is just pool quality. They don't make how this thing will behave clear to the user, or give control. They fail at anything that needs an abstraction or model, not just APIs and shell scripts glued together. And "just ask AI" seems to be the default fix.
That vibe coding they brag about as if it was a good thing, it shows.
Take their notation for describing permissions. The docs are not comprehensive, and in practice it doesn't quite work how they describe it.
Or their management of sub-agents. I once lost a sub-agent, it finished and disappeared from UI. Apparently, you can't bring it back yourself: you have to ask the parent agent to do it for you. But the parent was Fable, and I ran out of credits, so I was locked out of using my opus sub-agent because of it.
Or an even more grotesque example: when you paste your claude API token to authorize, it covers characters with *. But it seems like an LLM has hallucinated a limit of API key length and the tail of your key stays visible.
What amazes me is how, for a vibe coded product where all they have to do is use their AI to fix things ... NOTHING EVER GETS FIXED!
I've probably gone to file 20 bugs. In all 20 cases there wasn't just one issue already filed for it: there were several, each which had a bunch of upvotes. And in all 20 cases ... every. last. one. ... Anthropic closed the ticket with no comment.
IF YOU ARE GOING TO HAVE A SHITTY VIBE CODED PRODUCT, AT LEAST USE YOUR SHITTY AI TO FIX THE SHITTY PROBLEMS!
I mean. This is what happens with vibe-coded projects. When there's no actual software engineering going on, I wouldn't expect anything better than this.
I suspect it must depend on how one manages their codebase - wrt to docs, ADRs, and general guardrails.
For me it is not great for design work - Fable is way better, and 4.8 was conservative and thus better (Opus 5 seems to jump to conclusions far more eagerly). But for overnight builds, where I give it 8hrs worth of work on LLDs created by Fable - its great. Where Opus 4.8 would often lose the plot and stop for questions clearly answered in the LLD - Opus 5 does manage to complete. Since it launched, I don't remember it ever disappointing me with builds. But designs? Boy, is this thing explosively stupid sometimes.
For me it's the opposite: I don't have $200 plus to throw at Anthropic every Month, and when I do get to use Fable it rips through my usage credits like there's absolutely no tomorrow.
Of course, the hilarious thing to me is that Anthropic likes to claim that the usage limits are because of resource allocation problems or something like that. Obviously no such issue exists, otherwise they wouldn't allow you to bypass it by just paying a bit more and it would be a hard limit. So usage credits are entirely their way of just screwing you out of more money.
Its tough to go from max account at home and pay per usage enterprise account at work with heavy usage limits... but the limits are there because pricing is insane. Feel like I'm in the $5 Uber rides phase at home.
API pricing is almost definitely profitable, but at this point I assume it's a small minority of their inference traffic compared to subscription usage, and unlikely to make up for the rest of their expenses on its own.
Meanwhile I can do all that and more with reasonix harness for Deepseek with a cache hit rate of 99%. And that's with unsubsidized American providers like cloudflare or Digital Ocean
People keep saying this but from what we’ve seen, Anthropic models are marginally profitable and earn back their costs over their lifetime. The company is burning money building the next versions and other ventures (e.g. verticals), but the models themselves have been profitable.
I think profitability is a matter of accounting. Inference is where money is made, but training is where money is spent. We keep getting new models every few months, but frankly the old models are still quite usable. I suspect labs will soon start specializing in expert models per use case so they can increase the lifespan of individual models, and change the profitability per model.
That's not the only reason to go to expert models. The more different domains you try to stuff in there, the more parameters the model needs to keep things coherent and not overload tokens in a way that induces errors. For example, if a model trained only on biology text sees "sonic hedgehog" there's no ambiguity, and this compounds for all the things that are "overloaded," in the training corpus, which turns out to be quite a bit.
yeah, i tried out GLM-5.2 when the news was all full of hype for that, and it's fine... definitely better value that API rates for claude. but comparing the value i got from that to the value i get from a claude max subscription... claude is way cheaper.
The $200 sub is customer acquisition cost to hook devs that then become the marketing team trying to get their company to bring in Claude (at the highly profitable API price).
They have something like 80% gross margins, are at a $100B/yr ARR, and are growing at 10x per year... If that keeps up, they're going to be doing more revenue than Google in a year ($400B ARR, 20% per year growth)
How can you sanely project the last 12 months forward? We have seen a huge uptick in usage. Last summer AI was a toy to most devs, now every enterprise developer I talked to uses it every day. Coding agent providers are surely going to hit market saturation in the near future.
Maybe, maybe not. Personally, I hope local AI eats their lunch so that the benefits are more decentralized and accrue more to society generally.
I don't think you're right about that last prediction, at all. And new use cases are opening up as these get smarter. I think things are going to get pretty weird.
But the point was that it really doesn't look like they're losing money on users, on average.
Where did that $100B figure come from? I thought they were at ~10B at the end of 2025, so they're either not at 100B yet, or they're growing way faster than 10x / year.
Good question, I heard it on a podcast, but going back to the transcript, looks like that's their forecast, not that they've hit it, they estimated a current $70B, but they've been revising their forecasts up, so yeah, it's probably >10x. Latest solid number they reported was $47B in May.
Amazon's first principle is the Customer Obsession. Making customers happy.
Fun bit is that the human psychology rates personal looking fixes better than having no issues at all.
For example, AWS overcharges you, you contact support, and more or less hassle free they refund or issue credits. The customer feels appreciated, or at least got something "extra" or "special treatment".
Meanwhile, any other (small) cloud. Simple, no weird charges. Even _most_ of network egress is free. But, no reason to call support or feel "extraordinary". Comes out as "meh" against Amazon's "top tier" support model...
Aws is an infrastructure company that builds services on top of that infra to sell more of it at a higher margin.
Anthropic trains models on AWS's (and GCPs, and Microslop's) infrastructure, then skims margin off of selling inference also on the infrastructure owned by the other companies.
Meanwhile they keep making _mistakes_ and announcing global token resets for everyone. (Like ones for Pro and Max users). Making them happy although Claude Code made them consume tokens faster or more than necessary in the first place.
I had same experience with OpenAI. I have the $200/month plan and use 5.6 Sol all the time. What would normally use about 2% of my weekly allowance burned through $100 of credits in 40 minutes.
Yeah, but now we can hire the Chinese instead for 1/100th the cost. It's an even better deal.
Plus we get to own, keep, run, do whatever with the model. We don't feel trapped. Moreover, it's something we can truly build on top of and own our own destiny.
Anthropic and OpenAI are the new Oracle (Oracle pre-AI; Oracle is even worse now). Expensive, feels like dealing with a lawyer, and not at all open. They just became infinitely less cool than they were a month ago.
The whole of our industry is going to migrate to open weights. We're smart enough to know this is the better deal and technical enough to be able to pull it off.
The only thing that might save these OpenAI and Anthropic in the near-term is an abundance of enterprise contracts negotiated with non-tech companies. They'll soak consulting firms and F500 companies for "AI" integrations.
Yeah, what bugs me about stuff like that is like they spend all this time and then they output several or minimal real testing to prove the theory It's like you're building your model to And just because it takes a long time to compute and do the testing, you'd rather publish your article and then try to get credit on something that hasn't really been proven. Look, prove your results. Study it. Ruggedize it. Make sure it works. Then, show us.
This stuff is fun lately, I've been working on an engine for a while and ported a little game over to the browser here. It's a voxel kart racer as a browser project. Think classic kart racing, but with a voxel aesthetic. It's a fun little experiment, so I figured I'd share it here!
reply